Design for Six Sigma
Many products and services that are developed encounter difficulties when they are launched. These issues can be costly to the business in terms of design iterations; reengineering; inspection and testing to identify defects and longer development times. The Design for Six Sigma approach aims to ensure that performance of the design is optimised and is supplied through capable well controlled processes.
Accountability
The challenge for an organisation to adopt a design for six sigma approach should not be underestimated. The design and development function needs to work in conjunction with suppliers and operations who know how their manufacturing processes perform and how they must be controlled. In order to prevent expensive design iterations and reengineering of the product resources and good disciplines must be invested in the design for six sigma approach. This requires leadership and all functions involved in design and development to be familiar with the approach and to speak the same language.
Define the Concept
Design for Six Sigma starts with identifying a need that a customer has for a product or service. Voice of the customer (VOC) techniques translate that need into a functional requirement that is critical to the satisfaction of the customer (CTS) and specification for that product or service via Quality Function Deployment (QFD). Once all the customer needs have been identified and the functional requirements (CTSs) have been specified design concepts that meets the requirements can be evaluated. TRIZ techniques help with developing these concepts.
Analyse the Design
Failure Modes and Effects Analysis (FMEA) enables the risks associated with potential failure modes of the design to be evaluted and prioritised and appropriate design and verification to be undertaken. In conjunction with the FMEA, measurement systems are evaluated and a product scorecard is compiled that evaluates process capabilities.
Optimise
Design of Experiments (DOE) can be carried out to enable the relationships between process settings, design features and performance of the design to be determined and the design to be optimised. These experiments include robustness experiments that ensure that the design performs satisfactorily across a range of process and operating conditions
Performance of the design can be predicted by Monte Carlo simulation. The simulation also allows changes to the design and process to be evaluated and optimisation experiments to be carried out.
Statistical Tolerancing techniques enable process and design specifications and tolerances to be determined that ensure that process and design capability requirements can be achieved.
Verify
Finally the design needs to be tested and verified and a control plan implemented for maintaining ongoing control and capability of the design and process.